This project's research activities officially ended in March 2021. Legacy in SmartAgriHubs Portal

Feed Supply Chain Management

Coordination

Xavier Xirgu

Team

Jaume Gelada

Feed Supply Chain Management

Optimise the animal feed supply chain and logistical efficiency through IoT-based silo stock monitoring.

Introduction

The animal feed industry, mainly represented by feed suppliers and livestock farmers, currently faces great inefficiencies due to outdated supply chain management. Stakeholders struggle with the timing and quantity evaluation when restocking their feed silos, significantly affecting cost and labour efficiency. This use case thus develops an integral feedstock management system to optimise the entire supply chain. Smart volumetric sensors and 3D cameras are used to accurately calculate the silos' stock levels at a very low cost. Since the IoT devices are powered by solar energy, they can be easily installed and function fully independent. Stakeholders can access the information through an application, helping them to use resources more efficiently, reduce logistical efforts and therewith benefitting the environment.

Automatised restocking of silos

In many industrial environments, bulk solid products are stored and processed in silos or tanks. Examples include grain and compound feed silos as well as tanks used in the batch processing of foods or minerals.

In these industries, one must be able to reliably determine the amount of content stored within a container at any given moment to provide an adequate supply and monitor the in/outflow of content. This determination may be made visually. However, in many instances, the container prevents any sort of visual determination of the present level of the contents. For example, many silos are composed of steel and/or concrete or other non-transparent materials and, therefore, any visual inspection of the content level of such containers would involve manually opening an aperture, which may pose hazards to personnel, and raise the potential of contamination of the contents.

This use case’s innovation innovation allows the remote monitoring of inventory levels in any kind of tanks, silos or bins. Initially, they deployed their INSYLO device in bins but quickly moved on to real farms with animal feed silos. Eventually they acquired two silos of their own, to make validating and testing improvements of the device much easier. At the end of the project phase, Jaume Gelada, the Use Case Coordinator, and his team manufactured and installed 375 devices on livestock farms in UK, Spain as well as Sweden and connected all of them to the cloud monitoring platform to give the farmers access to the information of feedstock in the silos.

The INSYLO device has a depth-sensing-camera powered by a solar panel, so it can function independently. The camera looks inside farm silos and collects data about the occupied space, the temperature and humidity within the silo. For this, state-of-the-art optical imaging and information communication is used. The use case has lately finetuned the manufacturing of the sensors, and the high-volume performance of the cloud as well as app. Not unusual for technological innovation, this has not happened without overcoming several hurdles. There was no off-the-shelf solution to monitor bulk solids in farm silos to rely upon. Therefore, the hardware had to be developed by the team. The development of such a product requires complex operations and planning, and a high level of detail and expertise. After overcoming these challenges, there is now a version that works reliably in harsh field conditions, such as extreme sun, snow, rain and dust. Through the continuous acquisition of key parameters, the solution helps farmers to reduce worktime while minimising queried urgent orders. In these regards, it also reduces logistics costs for the supplier.

To that end, the use case’s innovation leverages a method for assessing the amount of content stored within a container. Since farm silos have a constant, known shape, defined by a 3D model on a given reference system and storing a given amount of content such as livestock food. Through the 3D sensor installed at the top of the silo with an inward orientation towards the content, the use case team acquires a depth map including the whole or a portion of a surface of the content. By using a computing unit operatively connected to the sensor, a 3D surface model – representing the surface of the observed area - and a 3D level model – representing the top level of the content - is created. The 3D surface model is computed based on the given reference system by processing said acquired depth map and using the given position, orientation and field of view of the 3D sensor. The 3D level model is computed by removing from the computed 3D surface model the points corresponding to the interior walls of the container. The innovation leverages a function that calculates the difference between intersection points of both models. Thereby, the use case team is able to accurately determine the volume of the content from a distance.

Valuable lessons learned

At the center of the use case technology is the cloud-connected monitoring platform. Over the time of the project the team developed new business intelligence features such as security and privacy tools as well as three business intelligence modules by UOC regarding the feed demand forecast, automatic restocking process and logistics optimisation.

These algorithms or modules provide multiparametric optimisation using real data from Spanish feed manufacturer Batallé. Exact knowledge of inventory levels in silos is critical for supply chain optimisation. When done properly, this should have financial impact in terms of a reduction in logistics and production costs by up to 30%. In terms of logistics, a reduced number of trips by feed trucks consequently results into environmental and social impact in terms of pollution, less traffic, noise and other inconvenience.

However, a dashboard or software can only show its entire benefits if all the functionalities are used correctly. Therefore, the use case team oversees each installation. Over the research period, they learned that only a real minor percentage of the devices could not be installed due to difficult or impossible access to the silo using the current means. To optimise the remote guidance of the installation process during the Covid-19 pandemic, the team prepared training content to the local installer in order to ensure the results and quality of the innovation. In addition to that, during the first installations INSYLO technicians also provided online videoconference-support to guide the installer. Luckily, the initial tests were carried out before the pandemic so the technicians already had a clear and deeper understanding of the hardware or software problems which might occur and how to solve them.

All new ventures and projects worth doing are challenging. One of the key insights that this use case has to offer is: find the right venture, the right project or the right problem worth solving and working for. One of the signs that you are on the right track is that the people you speak to care about the problem in their day-to-day work. They give you time and direction to arrive at solutions, even after having tried many alternative methods to solve the same problem. These are your champions, the real advocates for the solution you build - and it is critical to ensure their feedback is considered.

Key performance indicators

The implementation of the IOFEED system has reduced drastically the workload produced by sales department on chasing orders from clients. The service shows direct impact on livestock feeding logistics.

First, clearing the activity of stock run-outs. An average run-out cost was estimated to be 112€, based on the first four months of 2019 cost of run outs. Orders are now being put in advance (2-3 days) with To-Be-Confirmed (TBC) status using IOFEED. This saves a gap on the truck as it has a positive effect on planning and it has provided a base to align with other orders, saving a lot of time. Everything indicates that having a larger sensor deployment would positively affect feed production, smoothing the production curve by forewarning the products and quantities to produce in a wider area.

The implementation at the Spanish partner contributed in a similar vein. In contrast to the UK pilot, Batallé Group operates with production contracts between them (the livestock owners and feed suppliers) and the farmers (the facility owners) to have the farmers raise the livestock on their farms. Hence, as the farmers are partly paid by a facility rental, any actuation on controlling each livestock growing period would help them maximise the profits. Irregular livestock feeding affects the growing rate. Therefore, suppressing the run-outs may affect positively the animal growth, avoiding ordering feed in excess reduces the feeding costs by avoiding inventory relocation tasks. Finally, as the UK pilot confirmed, workforce dedicated to chase farmers is dramatically reduced. Additionally, they pointed out their interest in exploring the opportunity to differentiate normal intake patterns from abnormal consumption patterns that may appear. This intake rate fluctuation may correlate with nonoptimal feed quality or anticipate an animal disease. Although direct benefits to the farmers might seem negligible, the opportunity cost of not having accurate inventory reports is huge. A wrongly reported inventory may produce an over-sized or not required order, which in turn means unnecessary costs. In order to comply with bio-security regulations, farm facilities must be kept empty and cleaned for a period of time. This means that every time a bin has more than one ton of feed not consumed before animal relocation to the slaughterhouse, this feed needs to be removed with specific equipment. The Spanish pilot has recorded an average of 20 operations per month with a cost of 175€ per actuation, which means an average cost of 30 000€ per year. These costs are likely to be saved by controlling the stocks accurately as the run-out costs measured at the UK pilot.

Batallé’s logistics are already highly optimised. With regards to delivery, having obtained a 2% improvement on their distance covered, showed the use case team how the proposed system behaves at a very high level of performance. Taking into account this percentage reduction in terms of distance covered, the researchers evaluated distinct scenarios where this distance reductions could be larger than the obtained logistics. Hence, assuming a realistic scenario where the system could deliver a 10% reduction in distances, enables an opportunity cost of 78 000€ for the same volume of logistics for 5193 deliveries to farms every year with an average load of 28 tonnes.

Smart Agri Hubs

Get in touch with the use case team via the SmartAgriHubs Innovation Portal

Go to portal

Achievements, products & services

Solar-powered, plug and play, affordable camera for silo monitoring

Providing forecasting and logistics optimisation for both the farmer and the feed supplier

3D mapping of the silo contents

Automatic alerts for ordering new feed

Use case partners

IOT Catalogue

EXPLORE THE IOF2020 PROJECT AND ITS COMPONENTS IN THE IOT CATALOGUE

Go to IOT catalogue